Paper-and-pencil vs. computerized administration of cognitive ability tests: a study addressing the question of equivalence

Computerized versions of several cognitive ability tests were developed for administration on the DELTA computer system (a multi-user system to support psychological assessment and counselling at the local employment offices). A study was conducted to examine the psychometric equivalence of some of these computerized tests and their paper-and-pencil counterparts, including the Standard Progressive Matrices (SPM), subtests of the Leistungsprufsystem (LPS) and subtests of the Berufswahltest (BWT). A counter-balanced design with repeated measures was used. With regard to the SPM and BWT subtests the cross-mode-correlations did not differ from the reliability coefficients of these tests and no significant medium-of-administration effects on test means were found. In contrast to these findings the equivalence to some of the LPS subtests seems to be doubtful (both unsatisfactory correlations and significant effects on test means were observed). Some reasons for the heterogeneous results are discussed in this at-tide.